Claim · model-written summary
Direct and derived harvest index approaches had comparable errors.
- Taxon
- Rice
Oryza sativat0 - Outcome
- errorse0
- Comparator
- direct and derived HI approachese0
Not reportedMaterialGeneCompoundOrgan / stageInteractorInteractor roleSecond interactorSecond interactor roleTreatmentSettingDirectionValueQualifier
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“Both direct and derived HI approaches yielded comparable errors”
t0“Rice”
Title
Sunil Kumar Jha; James Brinkhoff; Andrew J. Robson; et al. Remote Sensing 2026. doi:10.3390/rs18111785
Other claims from this paper (9)
- Method / resource March balanced forecast accuracy and lead time, with little later error reduction.
- Method / resource TabPFN had the lowest yield prediction RMSE of 1.85 t ha−1.
- Method / resource Ridge had the lowest biomass prediction RMSE of 3.05 t ha−1.
- Method / resource Tree-based models had the lowest derived harvest index RMSE, about 0.07.
- Method / resource Direct harvest index prediction had weak regional correlation of about 0.24.
- Method / resource Derived harvest index prediction had regional correlation near zero.
- Association Within site-seasons, harvest index correlated strongly with vegetation indices.
- Observation The harvest index and vegetation index relationships varied greatly across site-seasons.
- Recommendation The authors support satellite yield and biomass forecasting for operational use.